All the illegal immigrants I know have fake social security numbers so they can get real jobs. That way, they actually pay money into social security, but since they're not citizens they won't get any benefits from it later in life. So they're dumping money into our system for us, as an immigrant tax.

They also never commit any crimes because they don't want to get deported. One was a prostitute to pay for college, but that's a victimless crime, heyooooooo.

Drunk Drivers rarely hit people.

The majority of Drunk Driving related events don't result in passengers or other cars don't get hit.

Usually they get pulled over or "make it home ok..." or hit a stationary object.

Therefore the "human shield" arguement doesn't really hold.

Since there is very very rarely going to be a chance where a drunk driver would hit one person... there is an even smaller chance he would hit a second person.

Keep in mind that not only does it have to possible hit 2 people, but you have to be the second in the order and not the first!

So said "human shield" would be like having a 2 inch by 2 inch plate somewhere on your body to protect you from a bunch of stray bullets fired wildly.

I think the human shield argument works. If they crash into somebody and it stops the car, they're no longer moving so they're not gonna kill you. This only works when they stop after their first kill. If they just keep going and killing people all night GTA style, then we're all still fucked.

That would only work if drunk driving ALWAYS resorted in hitting people, or even did most of the time. 

Think of it this way.   Say your playing drunk driving dungeons and dragons.  Good old DDD&D.

Say you and an immigrant each have a 1 in 20chance of getting hit by a drunk driver... by rolling a D20.

The chances you both are going to rol that critical miss are 1/400.

So then there are 3 drunk drivers... two regular, one an immigrant.

Your chance of rolling that critical miss are now 3/20 - 1/400... so  = 59/400.

Now say you have a 1 in 20 chance being hit by a drunk driver.  and there are two on the street.

2/20 = 40/400.

EXCEPT the number is even worse then before with the first number, since you would have to roll initative vs the illegal immigrant to see who was out first... and going to be hit by the Drunk Driver.  So it'd really be more like 119/800 vs 80/800.

(1) The odds of rolling a given number on one of three 20-sided dice are not 3 in 20.

(2) [edit:  double checking something.  I'm kind of distracted right now, so it may take a while, but I'm taking down the questionable assertion in the meantime.]

I'm sorry. I have no idea what you guys are talking about anymore. I'm just not that flavor of nerd. Dice freak me out.

No, you seemed to miss the point.

There are 3 cases of drunk driving.  Hence 3/20.

While in the other case there are 2 cases of drunk driving 2/20.

Drunk driving and many others are "victimless" crimes unless they hit a victim.  But are still crimes.

So if there were 20 drunk drivers, each with a 1/20 chance to hit me, then I would have a 100% chance to get hit?  Or over 100% chance with 21?

http://www.edcollins.com/backgammon/diceprob.htm

"Now say you have a 1 in 20 chance being hit by a drunk driver.  and there are two on the street.
"2/20 = 40/400."

This is completely wrong. 

I did screw up the basic math, cause i was rushed before work....

but still... the point remains solid.

If you have a 1 in 20 to get hit... and the immigrant has a 1 in 20 of getting hit... there is still a smaller chance of being saved by a "Human shield" then there is actually getting hit again.

But if the crime rate for an untargeted crime is lower, then the risk of any individual becoming the victim of that untargeted crime is by definition less, no? So it seems to me that the only question is where your math is misleading, not if.

How is this logic mistaken?